SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912394021699584 |
|---|---|
| author | Liu, Zhining Amjad, Rana Ali Adkathimar, Ravinarayana Wei, Tianxin Tong, Hanghang |
| author_facet | Liu, Zhining Amjad, Rana Ali Adkathimar, Ravinarayana Wei, Tianxin Tong, Hanghang |
| contents | Providing Language Models (LMs) with relevant evidence in the context (either via retrieval or user-provided) can significantly improve their ability to provide better-grounded responses. However, recent studies have found that LMs often struggle to fully comprehend and utilize key evidence from the context, especially when it contains noise and irrelevant information, an issue common in real-world scenarios. To address this, we propose SelfElicit, an inference-time approach that helps LMs focus on key contextual evidence through self-guided explicit highlighting. By leveraging the inherent evidence-finding capabilities of LMs using the attention scores of deeper layers, our method automatically identifies and emphasizes key evidence within the input context, facilitating more accurate and grounded responses without additional training or iterative prompting. We demonstrate that SelfElicit brings consistent and significant improvement on multiple evidence-based QA tasks for various LM families while maintaining computational efficiency. Our code and documentation are available at https://github.com/ZhiningLiu1998/SelfElicit. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_08767 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence Liu, Zhining Amjad, Rana Ali Adkathimar, Ravinarayana Wei, Tianxin Tong, Hanghang Computation and Language Artificial Intelligence Providing Language Models (LMs) with relevant evidence in the context (either via retrieval or user-provided) can significantly improve their ability to provide better-grounded responses. However, recent studies have found that LMs often struggle to fully comprehend and utilize key evidence from the context, especially when it contains noise and irrelevant information, an issue common in real-world scenarios. To address this, we propose SelfElicit, an inference-time approach that helps LMs focus on key contextual evidence through self-guided explicit highlighting. By leveraging the inherent evidence-finding capabilities of LMs using the attention scores of deeper layers, our method automatically identifies and emphasizes key evidence within the input context, facilitating more accurate and grounded responses without additional training or iterative prompting. We demonstrate that SelfElicit brings consistent and significant improvement on multiple evidence-based QA tasks for various LM families while maintaining computational efficiency. Our code and documentation are available at https://github.com/ZhiningLiu1998/SelfElicit. |
| title | SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2502.08767 |